MIMO Transmission with Rank Adaptation for Multi-Gigabit 60 GHz Wireless

1. A wireless system comprising: a transmitter with a baseband processor responsive to groups of transmitter antenna arrays for communicating over directional beams; and a receiver with a baseband processor responsive to groups of receiver antenna arrays for communicating with said transmitter over said directional beams, said receiver including both a rank adaptation providing a transmit mode feedback to said transmitter and a blind beamforming providing a transmit beamformer index feedback to said transmitter and receiver groups of antenna arrays; wherein said blind beamforming comprises generating enhanced perturbation vectors, an iteration to generate new beamformers responsive to said perturbation vectors and a current beam beamformer, and determining a received power for all combination of transmitter and receiver antenna beamformers responsive to said new beamformers.

2. The system of claim 1 , wherein said blind beamforming comprises a stochastic gradient algorithm based blind beam forming.

3. The system of claim 1 , wherein said blind beamforming comprises determining optimal transmit and receive beamformers for updating said current beamformer.

4. The system of claim 1 , wherein said rank adaptation is responsive to a composite channel estimation with reduced complexity and higher accuracy from improved signal-to-noise-ratio contributed by joint effect of transmitter-receiver beamforming.

5. A method for wireless communication comprising the steps of: communicating over directional beams by a transmitter with a baseband processor responsive to groups of transmitter antenna arrays for; and communicating with said transmitter over said directional beams by a receiver with a baseband processor responsive to groups of receiver antenna arrays, said receiver including both a rank adaptation providing a transmit mode feedback to said transmitter and a blind beamforming providing a transmit beamformer index feedback to said transmitter and receiver groups of antenna arrays; wherein said blind beamforming comprises generating enhanced perturbation vectors, an iteration to generate new beamformers responsive to said perturbation vectors and a current beam beamformer, and determining a received power for all combination of transmitter and receiver antenna beamformers responsive to said new beamformers.

6. The method of claim 5 , wherein said blind beamforming comprises a stochastic gradient algorithm based blind beam forming.

7. The method of claim 5 , wherein said blind beamforming comprises determining optimal transmit and receive beamformers for updating said current beamformer.

8. The method of claim 5 , wherein said rank adaptation is responsive to a composite channel estimation with reduced complexity and higher accuracy from improved signal-to-noise-ratio contributed by joint effect of transmitter-receiver beamforming.

9. The method of claim 5 , wherein said rank adaptation comprises a capacity evaluation responsive to a composite channel gain obtained at said receiver, said transmission being adaptively chosen from one of a high rank spatial multiplexing and a rank-1 beamforming whichever gives higher throughput.

10. The method of claim 9 , where said capacity for said high rank spatial multiplexing and said rank1 beamforming is respectively given by C rankM = log 2 ⁡ ( det ⁡ ( I + P M ⁢ H ~ ⁢ H ~ ′ ) ) and C rank1 =log 2 (1+P|ũ{tilde over (H)}{tilde over (w)}′| 2 )), where {tilde over (H)} is a composite channel gain, P is the total transmit power, M is number of subarrays of antenna arrays.

11. The method of claim 10 , wherein said optimal transmission follows from the relationship C OPT =max(C rankM , C rank1 ).

12. The method of claim 10 , wherein said optimal capacity for said high rank spatial multiplexing comprises a spectral efficiency evaluation based on a practical receiver filtering responsive to a composite channel gain obtained at said receiver, said transmission being chosen from one of a high rank spatial multiplexing and a rank-1 beamforming whichever gives a higher throughput.

13. The method of claim 12 , where said practical receiver filtering comprises one of a minimum mean square error (MMSE) receiver, a zero-forcing receiver, or any other practical receivers.

14. A wireless system comprising: a transmitter with a baseband processor responsive to groups of transmitter antenna arrays for communicating over directional beams; and a receiver with a baseband processor responsive to groups of receiver antenna arrays for communicating with said transmitter over said directional beams, said receiver including both a rank adaptation providing a transmit mode feedback to said transmitter and a blind beamforming providing a transmit beamformer index feedback to said transmitter and receiver groups of antenna arrays; wherein said rank adaptation comprises a capacity evaluation responsive to at least one of composite channel gain and capacity prediction obtained at said receiver, said transmission being adaptively chosen from one of a high rank spatial multiplexing and a rank-1 beamforming whichever gives higher throughput; and where said capacity for said high rank spatial multiplexing and said rank1 beamforming is respectively given by C rankM = log 2 ⁡ ( det ⁡ ( I + P M ⁢ H ~ ⁢ H ~ ′ ) ) and C rank1 =log 2 (1+P|ũ{tilde over (H)}{tilde over (w)}| 2 )), where {tilde over (H)} is a composite channel gain, P is the total transmit power, M is number of subarrays of antenna arrays.

15. The system of claim 14 , wherein said optimal transmission follows from the relationship C OPT =max(C rankM ,C rank1 ).

16. The system of claim 14 , wherein said optimal capacity for said high rank spatial multiplexing comprises a spectral efficiency evaluation based on a practical receiver filtering responsive to a composite channel gain obtained at said receiver, said transmission being chosen from one of a high rank spatial multiplexing and a rank-1 beamforming whichever gives a higher throughput.

17. The system of claim 16 , where said practical receiver filtering comprises one of a minimum mean square error (MMSE) receiver, a zero-forcing receiver, or any other practical receivers.